CRB False Witness
Around 33 centuries ago, one of the most famous men in history hiked up a mountain probably now known as Jabal al Lawz in today's northwestern Saudi Arabia. There Moses met with God. God Himself carved commandments into stone tablets for Moses to share with His people, the Israelites. These commandments eventually became a major part of the legal foundation for western civilization.
One of these commandments preserved in the book of Exodus is "You shall not bear false witness against your neighbor." While most obviously commanding us not to lie, I believe this commandment goes well beyond lying. It probably also includes presenting true information in such a way that it will likely mislead when interpreted. A modern word that comes to mind along these lines is "nuancing".
Sadly the financial markets are full of this kind of thing. Charts, with their wealth of information, are one of the easiest ways to intentionally mislead others. Depending on the analyst's selection of data to use, time period to cover, axis type of chart (linear or logarithmic), and vertical axis span, the obvious interpretation of a chart can vary radically. It is not hard to present true information yet know it will be misinterpreted.
I have a personal anecdote on this. When I was around 13 years old or so, one of our national Senators came to speak at my school. This was well before the PowerPoint days so he presented his charts on big sheets of paper on easels. The Senator showed us one chart with a very sharp rate of increase. After he finished, I walked down to the front of the gym to take a closer look. It turns out his vertical axis wasn't zeroed.
What had looked like a mammoth 500% increase from the crowd was probably less than 10% when the tiny-labeled non-zeroed vertical axis was considered. In my young self-righteous fury, I actually wrote this Senator a letter chastising him for his misleading chart. His office even answered me. It was the first time in my life, and the last time, that I ever bothered writing to a politician. It's as useful as talking to a turnip.
At Zeal we love charts since they offer such an awesome and unparalleled perspective. We have already custom built thousands of different charts over the years and I am looking forward to personally building thousands more. But every time I create a chart, I try to carefully consider how it will likely be interpreted. I want my charts to accurately reflect and illuminate the particular market I am researching.
Unfortunately today a terribly misleading chart is tainting perceptions of the mighty global commodities bull. I cannot count the times I have seen analysts and investors use and misinterpret it. Most of these misinterpretations are unintentional, due to simple naiveté. But disturbingly I have also seen it used to intentionally mislead when an analyst knows better. This chart bearing false witness is a big problem.
If it was an obscure chart few investors considered, I wouldn't care all that much. But unfortunately this false-witness chart happens to be of the famous CRB index. The CRB index, of course, has been the flagship commodities index for decades. Having the CRB mislead on commodities' true bull-market progress is as appalling as if the NASDAQ failed to reflect technology stocks' true performance.
The reason CRB charts are bearing false witness today is because this index's new custodians radically changed its composition back in July 2005. The CRB's traditional equal weighting and geometric averaging among its component commodities were trashed. This tenth revision of the CRB created a new version of this index unlike any before in history. I wrote an essay back then explaining all of this.
The unprecedented tenth-revision CRB, or CRBr10, is utterly dominated by energy. Energy comprises 39.0% of this new index compared to 17.6% in the ninth-rev CRB, the CRBr9. So when oil corrected in late 2006, the new CRBr10 plummeted. The CRB's total breakdown led many analysts, including long-time contrarians, to wrongly conclude this commodities bull was over. I wrote another essay in January 2007 pointing out how horribly flawed it was to assume the CRBr10 and CRBr9 were comparable.
After that essay, some traders graciously wrote in to tell me about the CCI, or Continuous Commodity Index. The CCI is really the traditional CRBr9 we are all used to. It was created to preserve the historical ninth-rev CRB interpretation of commodities' performance in this new CRBr10 era. In February 2007, I wrote an essay on the CCI. I was so thankful and relieved to learn of its existence because it restored the accuracy and comparability of the traditional flagship CRB index.
Now a year later it is very disappointing to see most analysts and investors continue to focus on the new CRBr10 rather than the old CRBr9 in the form of the CCI. The CRBr10 is fine when viewed in isolation only since its July 2005 birth. But when it is grafted on to the old historical CRBr9, it bears false witness. Almost no charts acknowledge this big break in continuity and hence their obvious interpretation is hopelessly flawed.
So this week I want to revisit the true CRBr9 in its new life as the CCI versus the CRBr10. Your perception of commodities performance in 2007, and their future prospects, will be very different depending on which CRB you ponder. While commodities continue to look rather anemic from the CRBr10 perspective, they look amazing from the true historical CRBr9 perspective.
In these charts, the CRB is rendered in blue. Up until July 2005, it is the ninth-revision CRB. After July 2005, the tenth-rev CRB is grafted in. This is the way virtually all analysts today present the CRB, with no note of the huge discontinuity. Meanwhile the new CCI, the comparable continuation of the CRBr9, is rendered in red. You have to consider it to truly understand the CRB's progress in historical context.
You really have to journey back in time to understand the devastating impact of the CRB's false witness. Between late 2001 and mid-2006, the CRB index was in an absolutely beautiful secular uptrend. Its support was rock solid and never failed. If an investor wanted to know how commodities were doing as a sector, all he had to do was check out a CRB chart. Due to the natural smoothing effect of the CRB's geometric averaging, you couldn't ask for a nicer and tighter uptrend.
Then stealthily in July 2005, the tenth revision of the CRB index since its 1957 birth went into effect. It started trading on July 12th, a quiet time of the year when the markets don't get a lot of attention. And there really weren't a lot of folks talking about this transition. I wrote a single essay on it, published on the quiet July 4th vacation week, and then got back to more tradable analysis. So not many traders knew the CRB had even changed.
And for the tiny fraction that did know, things didn't look all that different initially. As this chart shows, the CRBr10 dutifully continued up within the CRBr9's tight secular uptrend for a year after the revision happened. So by the time mid-2006 arrived, virtually everyone in this sector had long forgotten about the obscure tenth revision of the CRB. Back then we all charted the CRB across its tenth revision as if it was perfectly historically contiguous.
Then in late-summer 2006, the oil price started plummeting. In the old CRBr9 days, this index wasn't heavily influenced by crude oil. Oil was only 5.9% of it by weight and the geometric averaging greatly smoothed out individual commodity impacts. But in the CRBr10 oil was suddenly its largest component by far with a massive 23.0% weighting, 4 times higher. And the traditional geometric averaging no longer existed to moderate its impact on the entire index.
For the first time in its entire bull, the CRB plummeted. It sliced through both its rock-solid support and key 200-day moving average like a hot knife through butter. Investors were shocked and terrified. Since everyone had long forgotten the CRB revision, they truly believed the same old historical CRB had violated its long secular support. The sky really was falling technically in CRB-land, so commodities Armageddon looked to be upon us.
This breakdown led to all kinds of very bearish theories on commodities. Even former commodities enthusiasts were falling all over themselves heralding the end of the commodities bull due to this high-profile CRB breakdown. But the problem was this breakdown was false. The CRBr10 was breaking down, but it wasn't comparable to anything historically, especially the CRBr9's uptrend. The standard trans-revision CRB chart, as presented and interpreted, was bearing false witness.
I again wrote about the new oil-dominated CRB in October 2006, trying to help investors see the real picture. But I felt like the little Dutch boy trying to plug the dike as it was hopeless trying to stop the flood of commodities despair. At that time, unfortunately I wasn't yet aware of the CCI so I had no CRBr9 to compare to the CRBr10 in my charts. So I had to ask investors to take it on faith that the old CRB would not have corrected anywhere near as hard with oil as this new CRB did.
Before that breakdown, the CRB carved its bull high in mid-2006 which was 99% above its early 2001 lows. Since then, the headline CRB has just ground sideways. It did rally in 2007, but as of the end of the year it still hadn't managed to eclipse its May 2006 high. To a technician, the headline CRB looks like it has just carved a massive double top ahead of a secular bear. Thankfully this is a false witness.
In these charts, the only truly comparable lines are the blue CRBr9 one up until the tenth CRB revision in mid-2005 and the red CCI one after. This is the only way the CRB index is constructed and calculated the same way to ensure perfect comparability. While the headline CRBr10 swooned, the historical CRBr9 was stronger than ever living on in the form of the CCI. The true CRBr9 just hit all-time nominal and bull highs!
In December, on fully seven separate trading days, the CCI ascended to new closing highs. They are the highest levels yet seen in this bull and the highest nominal levels ever. But if you adjust the CRB for inflation as is necessary and prudent over multi-decade timespans, this index would have to soar well over 1000 today to hit a new all-time real high.
Thus bull to date, the truly comparable CRBr9 was up 160% as of late last month! And technically-oriented traders should note that its secular support was never broken! Not only did the CRB not break support, but its old secular resistance actually became new higher support as the CRB broke out of its secular uptrend in early 2006 ahead of oil's sharp correction. And that infamous oil correction really didn't even faze the CRBr9 due to oil's modest equal weighting and the old-school geometric averaging.
So as you can see, your interpretation of late 2006 and 2007 in commodities is radically different depending on whether you consider the new CRBr10 or the traditional CRBr9. Either way, the CRBr10 is simply not comparable to the CRBr9 in past years. So even if you like the CRBr10, and it does have its merits, it is illogical and misleading to graft it on to the CRBr9 with no explanation. Hence I propose big warning symbols on all trans-2005 CRB charts to stop them from bearing false witness and misleading investors.
This vast gulf between ninth-rev and tenth-rev CRB performance is even more interesting when viewed just since the tenth revision. The traditional CRBr9 in the form of the CCI held pretty tight with the new CRBr10 for its first quarter of existence, but then the CCI started to pull away. This gap has only continued to widen since. Actually this is rather curious considering the CRBr10's huge oil weighting and oil's massive upleg in 2007.
Since the day the CRBr10 went live, it is only up 15%. Meanwhile the CRBr9 in CCI form has soared 52%! There is really no comparison between the CRBr9 and the CRBr10. In your mind, carefully consider the blue CRBr10 line in isolation and then the red CRBr9 line in isolation. Your interpretation of this commodities bull's performance of late will vary tremendously based on which CRB you are pondering.
One has broken its support and 200dma, and is just now clawing back up to what really looks like a secular double top. In pure technical terms, this is not a market I'd want to invest in. Meanwhile the other one looks incredibly bullish. It is climbing ever higher in a beautiful uptrend while periodically bouncing off both its support and 200dma like clockwork. And its rate of ascent is not extreme so it certainly looks like a healthy secular bull, not an out-of-control bubble.
In CRBr10 terms, 2007 was a strong year but no new highs were made and commodities were merely recovering to a potential double top. But in CRBr9 terms, 2007 was an amazing year with consistent new highs emerging within a healthy unfolding secular bull. There was nothing at all to be concerned about technically and the uptrend points to continuing higher prices ahead.
So whether you are a commodities investor, speculator, or analyst, please realize that today's CRB is nothing at all like the historical one you remember from the first half of this decade. Today's CRB is only comparable back to July 2005, and it is utterly dominated by oil. In a very real sense, the CRBr10 is largely an oil proxy. It really doesn't reflect other commodities' progress all that well.
If you are trading commodities or commodities stocks, know that any CRB chart that stretches back past mid-2005 is not comparable. It compares apples to oranges and is totally useless across the tenth revision. So any time you see technical CRB conclusions reached across mid-2005, know they are nonsensical. At best the person reaching the conclusions is naïve and at worst he is intentionally misleading you. Any long-term CRB chart that mixes the CRBr9 and CRBr10 bears false witness to this commodities bull's true progress.
If you are an analyst, and are blessed to be in a position to influence traders, for heaven's sake please don't use long-term CRB charts to make points! Use the CCI instead, which is perfectly comparable from today all the way back to the ninth CRB revision in late 1995. If you are using trans-2005 CRB charts to make technical points, and your readers read an essay like this one, you will instantly and irrevocably lose lots of credibility. And in this business credibility is everything. We have to tell the truth, always, and never bear false witness.
Any comparison of the CRB across its radical tenth revision in July 2005 is hopelessly flawed. The CRBr10 is nothing like the nine CRB revisions that came before it. It is an entirely new beast altogether. Now the CRBr10's calculation methodology is valid, and superior in some ways, but it is simply not comparable with the past CRB. If you want an accurate and true reflection of this commodities bull's progress, you have to use the true historical ninth-rev CRB now known as the CCI.
So as we head into 2008, please realize that today's tenth-rev CRB is essentially just a proxy for oil. Oil's swings utterly dominate this index. When oil corrects soon here as it ought to, the CRB is going to get pummeled down hard. It will indeed look like a double top on a long-term chart. And the commodities Chicken Littles will come out of the woodwork boldly proclaiming that "the sky is falling!" Don't believe it.
Instead look to the old-school Continuous Commodity Index, where the historical ninth-rev CRB we all know and love lives on in new form. An oil correction will weigh on it too, but the CCI will probably only get dragged back down to support at worst. I am all but certain that even the most wicked oil correction you can imagine won't be enough to cause the CCI to fall under its support and break its uptrend. With oil's traditional equal 5.9% weighting and geometric averaging, it just can't do that much index damage.
At Zeal we live to study and trade the markets, so we really pay close attention to stuff like the tenth CRB revision that can really trip up investors who aren't aware of it. As students of the markets we constantly strive to deeply understand them. We have to be accurate and truthful in our analysis and we don't want to mislead anyone. So we won't foist off some long-term chart as comparable if it is not in reality.
Even before I learned of the wonderful CCI's existence, we remained big commodities bulls through the entire CRBr10 breakdown. From our research we knew it was just a temporary oil thing. And we continue to be very bullish today, buying elite commodities stocks to amplify the gains in their underlying commodities. Despite periodic bearish analyses claiming the contrary, this commodities bull is very much alive and well, thriving really, today. Fortunes continue to be won.
We just published the popular new January 2008 issue of our acclaimed monthly newsletter that has our outlook for key commodities sectors in early 2008. Subscribe today to get our latest cutting-edge analysis and see the actual real-world trades we are making based on all our research. First-time e-mail-PDF-edition subscribers will even receive a complimentary copy of this new newsletter, with your formal subscription starting next month.
The bottom line is the flagship CRB commodities index, if charted across its tenth revision in mid-2005, bears false witness. It misleads traders into assuming that the ninth-rev and tenth-rev CRBs are comparable even though nothing could be farther from the truth. The only truly comparable CRB is the CCI, which continues the ninth-revision CRB's components, weighting, and calculation methodology.
If you see someone trying to draw technical CRB conclusions across mid-2005 without using the CCI, know you are being misled. Odds are the analyst is merely naïve, but in some cases the false witness may be intentional to further the analyst's own hidden agenda. You can't thrive and trade optimally during a secular bull without a comparable and accurate yardstick to measure its entire bull-to-date performance.
Adam Hamilton, CPA
January 4, 2008
Thoughts, comments, or flames? Fire away at firstname.lastname@example.org. Due to my staggering and perpetually increasing e-mail load, I regret that I am not able to respond to comments personally. I will read all messages though and really appreciate your feedback!